Magnetic and structural properties of pure hematite

1 downloads 0 Views 597KB Size Report
program named MAUD [14], which combines the Rietveld ... 1 and 2 show the XRD patterns for the samples .... The authors are grateful to CODI (sustainability.
ARTICLE IN PRESS

Physica B 389 (2007) 145–149 www.elsevier.com/locate/physb

Magnetic and structural properties of pure hematite submitted to mechanical milling in air and ethanol L.C. Sa´ncheza, J.D. Arboledaa, C. Saragovib, R.D. Zyslerc, C.A. Barreroa, a

Grupo de Estado So´lido, Sede de Investigacio´n Universitaria, Universidad de Antioquia A A 1226, Medellı´n, Colombia Departamento de Fı´sica, Comisio´n Nacional de Energı´a Ato´mica (CNEA), Centro Ato´mico Constituyentes, Av General Paz 1499, 1650, San Martı´n, Buenos Aires, Argentina c Centro Ato´mico Bariloche (CNEA), 8400 SC de Bariloche, RN, Argentina

b

Abstract The mechanochemical treatment of a-Fe2O3 (hematite) powder in air and in ethanol at different times using a planetary ball mill was performed. X-ray diffraction (XRD) using Rietveld analysis and Mo¨ssbauer spectrometry showed partial transformation from a-Fe2O3 to a spinel phase for prolonged milling times in air, while only the a-Fe2O3 phase was obtained at all milling times using ethanol. Best XRD and Mo¨ssbauer fits were obtained by introducing two hematite components, one of them assigned to the precursor sample, and the other one to a nanoestructured hematite. Superparamagnetic hematites were only detected in the samples milled in air. An expansion of both a and c lattice parameters for the samples milled at all times in both environments were observed. Average grain sizes for the nanostructured hematites milled in air were always lower than those for the samples milled in ethanol at all times, reaching a minimum of about 11 nm. r 2006 Elsevier B.V. All rights reserved. PACS: 75.50.Tt; 61.10.Nz; 61.18.Fs; 61.82.Rx Keywords: Hematite; Mo¨ssbauer spectrometry; Planetary ball milling; X-ray diffraction

1. Introduction Hematite, a-Fe2O3, is an interesting iron oxide, which has been subject of intensive research for a long time [1–4]. The crystal structure is the same as that of corundum, Al2O3, which can be described as rhombohedral or ¯ or D63d, respectively. hexagonal with the space groups R3c Hematite orders weakly ferromagnetically (WF phase) below a Ne´el temperature TN of about 953 K. In this WF state, the spins lies in the basal plane (1 1 1) (rhombohedral unit cell) slightly canted from antiparallel alignment. The magnetic and structural properties of hematite are known to be affected by particle size [5–7], degree of crystallinity [8], pressure [9] and doping [1–4], among other things. The mechanical alloying of hematite has been recently reported in several works [5–7,10–13]. It is found that depending Corresponding author. Tel.: +057 4 210 56 30; fax: +057 4 233 01 20.

E-mail address: cbarrero@fisica.udea.edu.co (C.A. Barrero). 0921-4526/$ - see front matter r 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.physb.2006.07.042

upon milling conditions, structural phase transformations of nanostructured hematite to maghemite, magnetite and/ or wu¨stite can be induced. Most of the reported works focus the attention to phase transformations, but very few ones concentrate upon the conditions of progressive grain size reduction down to nanometer dimensions without inducing phase changes. In this work, the conditions for production of a-Fe2O3 nanoparticles under varying environmental and milling times are addressed. A detailed analysis of the crystallographic and some magnetic aspects of the milled products are also included. 2. Experimental a-Fe2O3 (Merck) powder with average particle size of about 4 mm and 99.9% of purity was milled. The mechanical milling was carried out in a planetary ball mill Fritsch Pulverisstte 5 by using 2 Cr-based stainless steel jars of 250 mL with 10 balls of 12 mm diameter made of the

ARTICLE IN PRESS 146

L.C. Sa´nchez et al. / Physica B 389 (2007) 145–149

same material. Different milling times were considered (0.5, 1, 3, 6, 12, 18 and 24 h) and the sample to balls weight ratio was fixed to 1:20. The powders were milled in air and ethanol (10 mL) both at atmospheric pressure. Each sample was milled at 390 rpm for intervals of 1 h and break periods of 30 min up to completion of the desired milling time. The jars were opened to the atmosphere during 2 min after each hour of milling for the samples milled in air. The crystallographic structure was investigated by X-ray diffraction (XRD) using a CoKa (l ¼ 1.78897 A˚). Data were collected in the 2y range of 20–801 with scanning step of 0.0141. Rietveld refinement was done by using a program named MAUD [14], which combines the Rietveld method and a Fourier transform analysis, well adapted especially in the presence of broadened Bragg peaks. The scale factor, sample displacement, incident intensity, unit cell parameters, four-order polynomial background, the average crystallite size, microstrain parameter and the relative volume fraction were refined. The average crystallite size was assumed to be isotropic, and no texture effect was considered in the present study. Mossbauer spectra (MS) were recorded at room temperature (RT) in the standard transmission geometry, using a Co57/Rh source. 3. Results and discussion Visual observation of the samples milled in air shows that the color changes gradually with increasing milling time from red (precursor sample, 0H) to reddish-brown (24H). On the contrary, the color of the samples milled with ethanol remains red independent of the milling time. Figs. 1 and 2 show the XRD patterns for the samples milled in air and ethanol during 0, 0.5, 1, 3 and 24 h (H).

Fig. 1. XRD patterns for the samples milled in air at different times. M accounts for a spinel phase.

Fig. 2. XRD patterns for the samples milled in ethanol at different times. To identify the pattern for a given sample, the milling time and its presentation order are the same as in Fig. 1.

All Bragg peaks of the XRD patterns for the samples milled in ethanol at all times were only assigned to the presence of a-Fe2O3. No other phase was identified. For samples milled in air a-Fe2O3 phase was only identified until 1H. However, from 3H on, new diffraction peaks located at 2y ffi 35:11, 50.71, and 52.31 were observed, the first two peaks were ascribed to the presence of a spinel phase and the third peak corresponded to the a-Fe phase. The latter phase is an impurity one originating from the jars and balls. The broadening of all peaks indicates the presence of nanoparticles [5,10]. The XRD pattern of the precursor sample (0H) was adjusted using Rietveld refinement by introducing only the crystallographic parameters of bulk hematite. However, the best fit for the other samples was obtained introducing two hematite phases corresponding one of them to the precursor sample (micrometer hematite) with fixed parameters as obtained for sample 0H, and the other one to the milled hematite (nanometric hematite). Of course for samples milled in air after 3H two additional phases (spinel phase and a-Fe) were introduced. Fig. 3 shows the average grain sizes of the nanometric a-Fe2O3 phases for all samples. A ball milling treatment in ethanol and in air for 0.5H is enough to reduce the average grain size of hematite from about 4 mm (0H) to about 41 and 21 nm, respectively. It is also noticed that the average grain size of the samples milled in ethanol gradually decreases from 41 to 19 nm with increasing milling times from 0.5 to 24 H, whereas it decreases from 21 to 10 nm in the case of the samples milled in air. Another interesting observation is that the average grain sizes of the samples milled in air were always lower than those of the samples milled in ethanol for all times. Moreover, for

ARTICLE IN PRESS L.C. Sa´nchez et al. / Physica B 389 (2007) 145–149

147

40

Average grain size (nm)

35 30 25 20 15 10 0

5

10 15 Time (hours)

20

25

Fig. 3. Average grain sizes of hematites milled at different times in air (O), and in ethanol (D).

Fig. 5. Fractional variation of the unit cell parameters for samples milled in ethanol. Solid spheres and gray stars are for Da/a and Dc/c, respectively.

Fig. 6. Relative phase fractions for different milling times in air. Fig. 4. Fractional variation of the unit cell parameters for samples milled in air. Solid spheres and gray stars are for Da/a and Dc/c, respectively.

the former it was possible to obtain hematites with grain sizes below about 11 nm, which are in the range of the critical grain sizes required to induce the phase transformation from hematite to a spinel phase [10] and also to obtain superparamagnetic hematite. Figs. 4 and 5 show the fractional variation of unit cell parameters in air and ethanol. The fractional variations of the unit cells a and c are defined as Da/a0 ¼ (aa0)/a0 and Dc/c0 ¼ (cc0)/c0, respectively, where a0 and c0 are the unit cell parameters for the precursor or micrometric hematite, whereas a and c are the unit cell parameters of the nanometric hematite [5]. Da/a0 and Dc/c0 were found to be positive for all milled samples indicating a lattice expansion. Moreover, the fractional variations of a and c exhibited the same tendency, ruling out the presence of anisotropic lattice expansion. In the case of the hematites milled in air, a maximum in the lattice expansion is

observed for the sample milled during 6H. In contrast, for the hematites milled in ethanol, the lattice expansion was the same for all milling times. The observed lattice variations can be attributed to the low grain sizes and the structural disorder of the hematites. In fact, it is well documented in the literature that lattice constant changes, contraction or expansion, are expected when the grain sizes decreases [15,16] as compared to the bulk values. On the other hand, the microstrain values for all samples ranged from 0.006 to 0.008. These low values are consistent with the lack of grain boundaries [10]. Fig. 6 shows the relative phase fraction for different milling times in air. A progressive increase up of 12H of the spinel phase with the respective reduction of a-Fe2O3 was observed. This spinel phase exhibited unit cell parameters of about a ¼ 0.838 nm. Figs. 7 and 8 shows the recorded MS for the precursor and the samples obtained for different milling times in air and in ethanol. The spectrum for the precursor sample was

ARTICLE IN PRESS 148

L.C. Sa´nchez et al. / Physica B 389 (2007) 145–149

Fig. 7. RT Mo¨ssbauer spectra for the pure and the samples obtained for different milling times in air.

Fig. 8. RT Mo¨ssbauer spectra for the sample obtained for 3H milling time in air and ethanol.

adjusted by introducing only one sextet with hyperfine field of 51.4(0.1) T, quadrupolar shift of 0.20(0.01) mm/s, and isomer shift of 0.38(0.01) mm/s relative to a-Fe. These

parameters clearly correspond to the values of the bulk or micrometric hematite. From 0.5H on, best fits were obtained by considering the occurrence of a hyperfine field distribution accounting for the broadening of the lines, which were associated to the nanometric hematite, and a well-defined crystalline component, associated to the micrometric hematite. The parameters for the crystalline component were kept fixed to the same values of the precursor hematite. This way of fitting is in accordance with the XRD analysis. For the MS of the samples milled in air after 3H, three additional components with distributed components were introduced. The first component with distributed magnetic hyperfine fields showing a maximum hyperfine field of 49.0 T, quadrupolar shift of 0 mm/s, and isomer shift of 0.32 mm/s was ascribed to the presence of a spinel phase. This refined isomer shift for the spinel component exclude the presence of a Fe2+ component, i.e. magnetite. We believe that maghemite instead of magnetite is the spinel phase being formed here because of the following reasons. It has been reported that the oxygen partial pressure has a critical influence upon the hematite to magnetite transformation, in such a way that in order to promote its formation, the bowl has not to be opened for a long enough time or alternatively, the milling should be performed in vacuum [12]. In our experiments, the bowls were opened after every hour of milling, and thus we allow fresh air to enter into the system. On the other hand, the colors of the powders were never black, but reddish-brown as mentioned before, also suggesting the presence of maghemite. According to the temperature– oxygen pressure diagram for the hematite–magnetite system presented by Zdujic et al. [12], our milling

ARTICLE IN PRESS L.C. Sa´nchez et al. / Physica B 389 (2007) 145–149

conditions do not favor the magnetite formation, because we are working above the line separating both regions (i.e. in the hematite region). The transformation of a-Fe2O3 to maghemite for the milling in air is due to the greater energy transferred to the powder for the mill, contrary to the milling in ethanol where this avoids the direct contact between powder and elements of the mill. The second component with distributed hyperfine fields was associated to iron coming from the jars and balls. The last component with distributed quadrupolar values suggested the presence of superparamagnetic hematite grains [4] with average sizes smaller than or equal to 8 nm. This component is a central doublet with quadrupole splitting ranging from 0.64 to 0.75 mm/s, isomer shift of about 0.36 mm/s and total area of about 371%. 4. Conclusions Significant structural changes of a-Fe2O3 powders submitted to mechanical milling in air and ethanol are reported. The present results show that nanostructured hematites without phase transformations are obtained after milling micrometric hematite powders in ethanol from 0.5 up to 24H and in air from 0.5 and 1H. From 3H on, partial transformation was obtained in air (12% approx.) of the precursor powder to a spinel phase in accordance with Rietveld refinement and MS, and explained like a crystallographic change relative to the mechanical energy ejected to the sample to reduce the grain size. In the milling with ethanol only nanoparticles and microparticles of a-Fe2O3 were detected. The magnetic hyperfine fields depend of the particles sizes and the broadening of the lines can be explained by using particle size distribution. In the case of the samples prepared in ethanol environment, no spinel and iron phases were detected at any milling time. In ethanol, the average grain size doubled the grain size of the hematites in air at all milling times.

149

Acknowledgments The authors are grateful to CODI (sustainability program for Solid State Group 2005–2006) and COLCIENCIAS (CIAM-2005 project and Excellence Centre of New Materials, contract RC-043 2005).

References [1] C.A. Barrero, J. Arpe, E. Sileo, L.C. Sa´nchez, R. Zysler, C. Saragovi, Physica B 354 (2004) 27. [2] C. Saragovi, J. Arpe, E. Sileo, R. Zysler, L.C. Sanchez, C.A. Barrero, Phys. Chem. Miner. 31 (2004) 625. [3] R.E. Vandenberghe, C.A. Barrero, E. De Grave, E. Van San, G.M. da Costa, Hyperfine Interact. 126 (2000) 247. [4] A.H. Morrish, Canted Antiferromagnetism: Hematite, World Scientific Publishing Company, Singapore, 1994. [5] S.J. Stewart, R.A. Borzi, E.D. Cabanillas, G. Punte, R. Mercader, J. Magn. Magn. Mater. 260 (2003) 447. [6] R.D. Zysler, D. Fiorani, A.M. Testa, L. Suber, E. Agostinelli, M. Godinho, Phys. Rev. B 68 (2003) 212408. [7] R.A. Borzi, S.J. Stewart, G. Punte, R. Mercader, M. VasquezMansilla, R.D. Zysler, E.D. Cabanillas, J. Magn. Magn. Mater. 205 (1999) 234. [8] M.Z. Dang, D.G. Rancourt, J.E. Dutrizac, G. Lamarche, R. Provencher, Hyperfine Interact. 117 (1998) 271. [9] C.L. Bruzzone, R. Ingalls, Phys. Rev. B 28 (1983) 2430. [10] N. Randrianantoandro, A.M. Mercier, M. Hervieu, J.M. Greneche, Mater. Lett. 47 (2001) 150. [11] S. Bid, A. Banerjee, S. Kumar, S.K. Pradhan, Udayan De, D. Banerjee, J. Alloys Comp. 326 (2001) 292. [12] M. Zdujic, C. Jovalekic, Lj. Karanovic, M. Mitric, D. Poleti, D. Skala, Mater. Sci. Eng. A 245 (1998) 109. [13] M. Hofmann, S.J. Campbell, W.A. Kaczmarek, S. Welzel, J. Alloys Comp. 348 (2003) 278. [14] The MAUD program, /http://www.ing.unitn.it/maud/S. [15] V.R. Palkar, P. Ayyub, S. Chattopadhyay, M.S. Multani, Phys. Rev. B 53 (1996) 2167. [16] Y. Duan, J. Li, Mater. Chem. Phys. 87 (2004) 452.